Database error ? using 4.4.1

Hi @mbsharp ,

Thank you for sending the log lines and for flagging that this started after the 4.4.1 update. That is exactly the right instinct after an update, and the detail you included made this quick to check.

I have good news and a bit of unwelcome news. The unwelcome part is that these are not NextGEN errors. The good news is that the real cause is something your host can fix quickly.

What these errors mean

Both lines contain the same MySQL code: “Errcode: 28 – No space left on device”. That message comes from the database server, not from a plugin. MySQL writes temporary working files to disk while it runs certain queries. When that disk fills up, any query that needs a temporary file fails.

That is what you are seeing. It is not a bad query and it is not a broken table. The database server ran out of room to work in.

Why NextGEN’s name is on both lines

WordPress prints the call stack of whatever query happened to fail, so the plugin that was running at that moment gets named. Two things make NextGEN a frequent name on that list, and neither is a fault.

First, NextGEN saves a small routing value at the end of each page load. That often makes it the first thing on the page to write to the “wp_options” table, so it is the first to hit the wall. The failing query in that line (“SHOW FULL COLUMNS FROM wp_options”) is WordPress core, not ours.

Second, the gallery query in your other line groups and sorts its results. MySQL always uses a temporary file for that kind of query, so it fails immediately whenever the disk is full.

Once the disk has room again, both messages stop. You will likely see the same errors from other plugins in the meantime.

What changed in 4.4.1

I checked the 4.4.1 code against 4.3.0 directly. The two files named in your log are byte for byte identical between the two releases, so nothing in this area changed.

There is one small connection worth mentioning, though. The 4.4.1 fix you noticed (“broken album and gallery links no longer fill the error log with database errors”) stopped those requests from failing early. They now run the full gallery query instead. On a server that is already out of disk space, that means a request which used to fail cheaply now asks MySQL for a temporary file and fails there instead. The symptom moved. The disk problem was already present.

What to ask your host

Please ask them to check three things:

1. Free disk space on the database server, and on the volume that holds the MySQL data directory.
2. Free space in the MySQL temporary directory, which is often on a small partition of its own.
3. Whether a quota on your account has been reached, if these sites are on shared or managed hosting.

Since you are seeing this on more than one site, a shared volume or a shared quota is the most likely explanation. That is worth mentioning to them, because it points away from any single site being at fault.

Two things worth a look on your side

Your log shows a gallery rendering inside a sidebar widget on your theme’s 404 page. That means every hit to a missing URL on that site builds a full gallery. If the site is getting crawled or scanned, that adds up fast, both in database writes and in cached gallery data. Removing the gallery widget from the 404 sidebar would cut that load a lot.

It is also worth checking the size of your WordPress debug log itself. On a site logging at this volume, the log file can be what fills the disk.

Nothing is needed from you in NextGEN right now, and there is no reason to roll back to an earlier version. The fix here is disk space on the server.

If your host confirms the disk was full and clears it, and you still see these errors afterward, please send me a fresh log along with the site’s cached data size and I will take another look. I am also happy to review your host’s reply if it is unclear.

Best,
Mihai